
A $20-tier Reel with three empty topics: an open-source SEO backend from a Tunisian developer, 950 stars in 53 days
What it is An open-source SEO backend you start with docker compose. It does site audits, rank tracking, keywords and backlinks. You bring your own paid keys such as DataForSEO, and there is also a cloud version starting at $49.
Why it took off On Sep 30, the Instagram repost account @gittrend.io (which advertises a $20 placement) posted a Reel with 81k views. It took the repo from about 100 stars to 600+ in two weeks. After that it steadily added a dozen-plus stars a day by claiming the semrush-alternative and screaming-frog topics that the big players had left empty.
BHow you could play it Pick a proven category of "open-source X alternatives", build the side the leader doesn't cover, claim the competitor-name topics the leader left unused, buy one repost-account short video in the first launch window, and prepare a second wave of exposure in advance.
- Data (updated daily) · Repo created 2026-09-22 · 988 stars · 7 days +52
GitHub · stars Data 2026-10-09 → 2026-11-20 (43 days, 42 observations). Points before 10-28 come from deeply / archive backfill, not every day has one
RankMeFast, a self-hosted SEO platform written by one person, started its repo in Sep 22, and by Nov 14 had 953 stars. It isn't the first in the category; the leader, OpenSEO, has 27.6k. Its stars came from distribution: an Instagram repost-account Reel, a move to claim empty GitHub topics, and a README that reads like a commercial product. Below: how much each of these mattered, and which parts you can copy and which you can't.
A Reel with 80k views
In Sep 30, @gittrend.io, an Instagram account that reposts GitHub projects (192k followers), posted a Reel. The caption opens: "RankMeFast is a self-hosted SEO platform that audits your site with 26 automated checks and explains every fix in plain words..." and ends with "More like this → gittrend.io". So far it has 81k views, 1,244 likes, 366 comments and 1,822 shares.
When the Reel went up, the repo was only 8 days old, with about 150 stars. It gained 72 that day, and 265 stars over that week (09-29 to 10-04) (daily star gains from Trendshift).
The account's bio links to gittrend.io/promote, where the rates are posted: one Reel is $100 at list price and $20 before Dec 1; one Story is $20. Whether RankMeFast paid can't be verified. Still, a repo 8 days old with 150 stars being organically picked off the trending charts by a 190k-follower account is unlikely (inference).
What it is: an SEO backend wrapped around data providers
The author, Selmi Abderrahim, is a full-stack developer from Tunisia with 161 followers on GitHub. He used to mostly write Django templates, and his best repo has 123 stars. RankMeFast has 125 commits in total, all by him.
The product is an SEO suite you start with docker compose up: site audits (26 checks, sorted into Fix now / Watch / Passed), rank tracking on Google / Bing / YouTube / Amazon, keywords, backlinks, Search Console and GA4 integration, brand visibility in AI answers, white-label reports, team seats, alerts, and an MCP server for AI assistants. The UI, emails and docs come in 7 languages, including Arabic and Chinese. License: AGPL-3.0.
It doesn't crawl pages or scrape Google itself. All data comes from keys the user supplies for DataForSEO, the Google APIs, or Firecrawl, and the project says it adds no markup. Each data source has a fake backend, so you can click through the whole product without any keys. Alongside it is an official cloud version at $49 / $159 / $499 per month across three tiers, marked as public beta.
How the stars came in: push, pause for two weeks, then a slow climb
| Timeline | How much it grew | What happened those days |
|---|---|---|
| 09-22 → 09-28 | About 100 | The author opened a dozen-plus issues and fixed them one by one. A few AI-written news and review sites indexed it the same day. |
| 09-29 → 10-04 | +265 | 09-30 @gittrend.io's Reel; 10-02 a 15k-subscriber YouTube channel posted an AI-generated intro, 215 views |
| 10-05 | +182, daily high | The cause isn't confirmed. The same day, RepoSquare published a hands-on long read, "Open-source Semrush alternatives," comparing it with OpenSEO. The author posted "539 stars on GitHub" on LinkedIn. |
| 10-06 → 10-13 | About +190 | Residual momentum; a Chinese mirror appeared on GitCode |
| 10-14 → 10-29 | About +45 | No new exposure, nearly flat |
| 10-30 → 11-14 | +172, 10–19 stars a day | No single spike; +7 over the last 100 days, +7 over the prior 63 days |
Source: daily star gains from Trendshift; after 10-25, daily star counts from GitHub.
The two quiet weeks in the middle show that the first wave was entirely driven by distribution. Once the Reel's buzz faded, the product itself didn't spread on its own. It never hit Hacker News, has no Reddit threads, and X has no discussion of it at all.
The slow climb of a dozen-plus stars a day afterward doesn't match any single event. It looks more like people who searched "open-source Semrush alternative" and found it. That's where topics come in.
Three topics nobody had claimed
It stuffed 17 terms into its GitHub topics, including semrush-alternative, semrush-free and screaming-frog. Sorted by stars on GitHub, 11-14:
- The topic "semrush-alternative" has 22 repos in total; it ranks first, and second place has 366 stars;
- The topic "screaming-frog" has 32 repos in total; it ranks first, and second place has 96 stars.
OpenSEO uses terms like seo-audit, seo-tools and keyword-research. It ranks first under seo-audit (677 repos) and seo-tools (1,566 repos), but it never touched competitor names like "semrush" or "screaming-frog". RankMeFast took the top spot on those pages with under 1,000 stars. Under a broad topic like seo-audit, it wouldn't rank near the top.
Further out are aggregator sites of AI-generated content and repo review sites, plus one-click deploy templates on Railway and RepoCloud that automatically pull newly created GitHub projects. Each brings in little traffic, but together they make up the long tail of a dozen-plus stars a day.
The category was validated by someone else
OpenSEO built the "open-source Semrush alternative" category first, back in late February 2026. By September 2026 it had about 20k stars, and it's now at 27.6k, still gaining 990 stars over the last 7 days. Plenty of people had already shown they would pay for DataForSEO themselves rather than pay Semrush $139 a month. When RankMeFast arrived, the search demand, listicle writers and aggregators were already there.
Compared head to head, the strengths and weaknesses are clear. The RepoSquare author installed both and ran them against his own site:
- RankMeFast needs 6 containers (MongoDB, PostgreSQL, Redis, plus API, worker and frontend). OpenSEO needs one. His first launch failed, and it only came up after he added an environment variable for the contact-form email.
- Without keys, RankMeFast's audit returned demo data. In his words: "It came back with one 'broken internal link' and 12 checks marked 'Not evaluated'. That broken link is not real." OpenSEO ships its own crawler, so you can audit your site without paying anyone.
- His verdict: "OpenSEO for a solo developer or one blog. RankMeFast for an agency that will pay for data anyway." Team seats, white-label reports and the plain-language "Fix now" layout are things OpenSEO doesn't plan to build.
So the product's breadth wasn't the source of the stars. What got repeated in the spread was only "self-hosted + 26 checks + plain-language explanations," and OpenSEO can say those too. Completeness did its job after the Reel brought people in: they opened the repo, saw the feature table, the 7-language badges, the 100% coverage badge and the cloud pricing page, decided it wasn't a toy, and clicked star.
Who else is doing it
Didn't find a clear copycat. The repos that popped up under the "Semrush alternative" search term in the same window (Sep 25 semrush-ai-tool at 366 stars, Oct 15 a repo called openseo at 49 stars) are riding the buzzword, not copying it. The SEO space now is topped by OpenSEO at 27.6k, claude-seo at 29k (a skill you install into Claude Code), and open-seo-mcp-skills at 9,500, with a tail of single-feature tools at a thousand or two stars: SerpBear, which only does rank tracking, and LibreCrawl, which only does crawling.
If it were you
B, worth trying. What you can copy is the playbook, not the category. One person, two months, one $20 Reel, and about 1,000 stars. The method is cheap and clear, and it can move to other "open-source X alternative" categories. The ceiling is visible too: its star count is 3% of the leader's, and growth stalled for two weeks after the first wave. Another open-source SEO dashboard would be fighting for topics that are already taken.
- Who it suits: one or two people who can get a product to the point where the feature list doesn't fit on one screen, who want a project with a thousand-plus stars on their résumé, and who are willing to spend a hundred or two hundred dollars on exposure.
- Who it doesn't suit: anyone aiming to build the next OpenSEO (that needs a genuinely different angle; RankMeFast shows "more complete" isn't enough), or anyone trying to squeeze into SEO as a category.
- How to do it: pick an "open-source X alternative" category where someone has already reached tens of thousands of stars, and take the side the leader skipped or chose not to take. RankMeFast targets teams and agencies. You could go lighter, switch ecosystems (a single Go binary, say), or aim at one type of user. Build a fake-data backend for every data source that needs a paid key, so people can click through the whole product without spending anything. Put a one-line positioning statement and a feature table on the first screen of the README, so someone can cut a video from it in 30 seconds.
- How to get stars: first look at which topics the leader used, and claim the remaining competitor names and "xxx-alternative" / "xxx-free" style terms. On those topic pages, the number two spot often has only tens to hundreds of stars. In the first two weeks after launch, find a GitHub repost account and get a short video posted (gittrend charged $20 before Dec 1). After the first wave, don't count on word of mouth. Have a second round of exposure ready (Show HN, the Reddit version of the same thing, a comparison post against the leader). Make the docs multilingual, and Chinese mirror sites and non-English aggregators will index them as a side effect.
- Before when: gittrend's $20 price was written with a Dec 1 deadline. Topic squatting has no window, but each category has only one slot to claim, first come, first served.
- What would show you misread it: ① RankMeFast's daily star gains drop to single digits over the next month, meaning the topic long tail can't hold up and this playbook is only worth one wave. ② Someone runs this playbook (repost-account Reel plus squatting on topics) for two months and stalls at two or three hundred stars, meaning the Sep 30 wave relied on luck or on a channel nobody has identified. The Oct 5 182 stars still have no known source. ③ GitHub starts limiting topic counts or cleaning out competitor-name topics unrelated to the project.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one every-app/open-seo(OpenSEO) Open source 2026-02-27 27,639 stars, 3,866 forks, +7 over the last 990 days; openseo.so's Tranco rank is around 270k and still climbing The category leader: MIT, single container, built-in crawler, auditable without any key, hosted version $10/month. RankMeFast has more modules (teams, white-label, 7 languages) but needs 6 containers, shows only demo data without keys, and its cloud version costs much more. AgriciDaniel/claude-seo Open source 2026-02-07 28,995 star A skill installed into Claude Code, with no self-hosted UI or database Ryze-AI-Adgent/open-seo-mcp-skills Open source 2026-08-29 9,490 star An MCP skill pack tied to Ryze's paid service, suspected of star inflation towfiqi/serpbear Open source 2022-11-24 2,184 star Rank tracking only PhialsBasement/LibreCrawl Open source 2025-09-27 1,114 star Crawling only (a Screaming Frog alternative), no paid API needed janreges/siteone-crawler / StJudeWasHere/seonaut Open source 2023-10 / 2022-03 about 1,007 / 871 stars Technical crawling and auditing, no keywords, backlinks, or rankings Semrush / Ahrefs / Screaming Frog Other Years The industry standard: Semrush from $139/month, Ahrefs Lite $129/month Collects its own data, closed source, priced per seat. RankMeFast is a self-hosted shell wrapped around data vendors. Google Search Console / Bing Webmaster Tools Big-company features Years Free, used by everyone Only first-party data from your own site. RankMeFast plugs into it and adds competitor and keyword data. - Where users come from
- Verified Instagram repost account Reel (@gittrend.io, 192k followers): published 09-30, 81k views, 1,244 likes, 366 comments, 1,822 shares (TikHub); +265 stars that week. The account's promo page lists a Reel at $100, current price $20. Whether RankMeFast paid is unknown.
- Verified GitHub topic squatting: topics include semrush-alternative, semrush-free, screaming-frog and 12 others. On those three topic pages it ranks first by stars. OpenSEO doesn't use these topics.
- Verified AI aggregator sites, review sites and comparison posts: Reveneau(09-28), itechguides, prlist, repoquarry, hysenlabs, gitdiagram, GitCode mirror, Railway/RepoCloud deploy templates; a Reposquare 10-05 long-form post lists it alongside OpenSEO as two open-source options
- Verified Author's LinkedIn: "539 stars on GitHub. I built RankMeFast because every SEO report..." Posted after the stars came in; engagement numbers aren't visible.
- Verified YouTube: an AI-generated intro video by Alex Hitt 10-02, 215 views, negligible impact.
- Inferred HN / Reddit / X / TikTok / Product Hunt: nothing found on Algolia, Brave site:reddit, x-radar (60 days), or TikHub.
Why SelmiAbderrahim/rankme.fast (RankMeFast) took off
Data as of 2026-11-14 (our GitHub database, Trendshift daily star gains, and the GitHub API). GitHub's stargazers endpoint now returns 404 for all repos (GraphQL totalCount is also 0), so we can't see who starred or when. Daily gains come from Trendshift.
Bottom line
RankMeFast is a late entrant in the "open-source Semrush alternative" category. OpenSEO had already validated demand in the first half of 2026 (about 20k stars in September, 27.6k now). RankMeFast built a self-hosted version that looks more complete and more like a commercial product. The first real wave came from a Sep 30 Reel posted by the Instagram repost account @gittrend.io (192k followers, 81k views, 1,822 shares). The account sells promotions, with the Reel now priced at $20 and originally $100, but whether RankMeFast paid is unknown. After that, GitHub topic squatting, a batch of AI-generated aggregator and review pages, and the search term "open-source Semrush alternative" kept it adding a dozen-plus stars a day. It was never on HN, has no Reddit threads, and zero posts in our X corpus.
What it is
- A self-hosted SEO platform: site audit (26 checks, sorted into Fix now / Watch / Passed), rank tracking for Google/Bing/YouTube/Amazon, keyword research, backlinks, Search Console/GA4, AI answer visibility (AI Overviews, LLM citations), white-label reports, team seats, alerts, and an MCP server. The UI, emails and docs are available in 7 languages (English, Arabic, French, German, Spanish, Russian, Chinese). Full-stack TypeScript;
docker compose upstarts 6 services (web/api/worker/mongo/postgres/redis). AGPL-3.0. (Source: README) - No data collected in-house: users bring their own keys for DataForSEO, Google APIs, Firecrawl and an LLM. The project says it adds no markup, and vendor responses are shared in a cache across users. Every data source has a
fakebackend, so the whole product can be clicked through without any keys. - There's an official cloud version: Starter $49, Pro $159, Agency $499 per month, plus Enterprise and pay-as-you-go add-on packs. Annual billing gets 20% off. Labeled public beta. (Source: rankme.fast/pricing)
- The author, Selmi Abderrahim, is a Tunisian full-stack developer with 161 followers on GitHub. He previously worked mostly on Django templates (top repo: 123 stars). 2026-08 He also built dockstash (30 stars), a self-hosted project in the same style. All 125 commits to rankme.fast were written by him alone.
Star growth timeline (Trendshift daily data + our database)
| Phase | Stars gained | Events that line up |
|---|---|---|
| 09-22 created → 09-28 | About 100 (Reveneau's GitHub API-based write-up says "7 stars in 105 days") | 09-27/28 The author opened a dozen-plus issues against himself and fixed them; a batch of AI news and review sites indexed it the same day (Reveneau 09-28, itechguides 09-28) |
| 09-29 → 10-04 | +265 (09-30 single-day +72) | 09-30 @gittrend.io posts a Reel: 81k views, 1,244 likes, 366 comments, 1,822 shares; 10-02 Alex Hitt (15k subscribers) posts an AI-generated YouTube intro, 215 views |
| 10-05 | +182 (single-day peak), Trendshift TypeScript chart rank #20 | The cause is unknown. The same day, Reposquare published a long piece, "Open-source Semrush alternatives," comparing OpenSEO and RankMeFast. The author pushed 62 commits that day and opened about 30 new issues, and posted "539 stars on GitHub" on LinkedIn (timing matches: cumulative about 540). |
| 10-06 → 10-13 | About +190 | Residual momentum. GitCode set up a Chinese mirror, and gitdiagram added it. 10-09 10-12 |
| 10-14 → 10-29 | Only about +45, nearly flat | No new exposure. 10-26 The author merged a "port cloud" PR, 10-28 released 0.2.0 |
| 10-30 → 11-14 | +172 (10–19 stars a day), +7 over the last 100 days, +7 over the prior 63 days | No single spike, just a long tail: topic pages, aggregator sites, comparison posts, and searches for "open source semrush alternative". Around 11-11, growai_edtech posted a Reel that got only 646 views; negligible. |
Our GitHub database: stars 947, gain_7d 100, prev_7d 63, gain_30d 208, status "accelerating", trending_days_any = 0 (our trending data only starts collecting from 10-28).
Competitors
| Name | Where | Launched | Scale | How it differs from RankMeFast |
|---|---|---|---|---|
| every-app/open-seo(OpenSEO) | Open source + hosted | 2026-02-27 | 27,639 stars, 3,866 forks; +7 over the last 990 days; openseo.so's Tranco rank moved from 346k at 10-29 to 273k at 11-12 (traffic database) | The category leader. MIT, single container, built-in crawler, site audit works without any keys; hosted version $10/month. RankMeFast has more modules (teams, white-label, alerts, local/app SEO, 7 languages), but needs 6 containers, shows only demo data without keys, and its cloud version costs 5–50× more. |
| AgriciDaniel/claude-seo | Open source (Claude Code skill) | 2026-02-07 | 28,995 star | Different form: a skill installed into Claude Code, with no UI or database of its own |
| Ryze-AI-Adgent/open-seo-mcp-skills | Open source (tied to a hosted service) | 2026-08-29 | 9,490 star | An SEO skill pack that requires Ryze's paid service. Our earlier research judged that it inflated stars and borrowed the open-seo name. |
| towfiqi/serpbear | Open source | 2022-11 | 2,184 stars, +13 in the last 7 days | Rank tracking only |
| PhialsBasement/LibreCrawl | Open source | 2025-09 | 1,114 star | Crawling only (a Screaming Frog alternative), no paid API needed |
| janreges/siteone-crawler | Open source | 2023-10 | About 1,007 stars | Technical crawler / QA, can run in CI |
| StJudeWasHere/seonaut | Open source | 2022-03 | 871 star | Crawling plus auditing, no keyword or backlink data |
| Semrush / Ahrefs / Screaming Frog | Commercial SaaS / desktop | Years | Industry standard | Semrush from $139/month, Ahrefs Lite $129/month (per Reposquare). They collect their own data; RankMeFast is just a shell on top. |
| Google Search Console / Bing Webmaster | Big-company features | Years | Free | Only first-party data from your own site, no competitor or keyword database |
Self-hosted SEO isn't empty: OpenSEO already owns the "open-source Semrush" mindshare, with 29× RankMeFast's stars. RankMeFast is fighting for second tier, about on par with single-purpose tools like SerpBear and LibreCrawl.
Growth tactics (evidence)
- Instagram repost account Reels (confirmed as the main driver, paid or not unclear): @gittrend.io has 192k followers, with gittrend.io/promote and linkpls (an auto-DM comment tool) in the bio. The Reel caption: "RankMeFast is a self-hosted SEO platform that audits your site with 26 automated checks and explains every fix in plain words... More like this → gittrend.io". Views and engagement are raw Instagram data from TikHub. gittrend's promo page lists prices openly: Reel $100 list, $20 before Dec 1; Story $20; the account claims 30 days and 12M views. When the Reel went up, the repo was only 8 days old with about 150 stars, which doesn't look like a natural pick from trending. Paid placement is quite possible, but there's no evidence.
- GitHub topic squatting (confirmed): topics include 12 terms like semrush-alternative, semrush-free, and screaming-frog. On the topic:semrush-alternative (22 repos), semrush-free, and screaming-frog (32 repos) pages it ranks first by stars. OpenSEO doesn't use these topics, so those slots are open.
- Long tail from AI aggregator and review sites (confirmed to exist; traffic impact unmeasurable): Reveneau (AI-written news), itechguides, prlist, positioniseverything, repoquarry, hysenlabs, explaingit, gitnova, gitdiagram, gitreverse, one-click deploy templates on Railway and RepoCloud, and a GitCode Chinese mirror. There's also a long hands-on review by Reposquare 10-05 that lists it as one of two "serious" open-source options and concludes in favor of OpenSEO.
- The author's own LinkedIn post (confirmed, small impact): "539 stars on GitHub. I built RankMeFast because every SEO report I opened felt like it was written for someone who already knew SEO..." It was posted after the star surge, and no engagement numbers are visible.
- YouTube (confirmed, negligible impact): an AI-generated video by Alex Hitt 10-02, with 215 views.
- Not found: HN (Algolia returns zero results), Reddit (Brave site:reddit.com search returns zero), X (x-radar shows zero posts in 60 days), TikTok (TikHub search for "rankmefast" returns zero), Product Hunt.
Why it's working
- The category is validated; it's a follower riding existing demand: OpenSEO launched in February and reached about 20k stars by September, which shows lots of people want to skip the Semrush subscription and pay DataForSEO themselves. When RankMeFast arrived, search and listicle traffic for this category already existed.
- A 80k-view repost Reel jump-started it: 09-30 it gained 265 stars that week, then 10-05 gained 182 more, pushing the count from 100 to over 600 in two weeks. After that there was no new exposure, and growth from 10-14 to 10-29 was nearly flat, which means this wave was entirely driven by distribution.
- The README and product look like a complete commercial product: a feature table, 26 audit checks, 7 languages, a 100% coverage badge, a demo mode that works without keys, Makefile ops commands, and a pricing page for the hosted cloud version. That's easy to pitch in the repost video's 28 seconds, and visitors who click through don't feel it's a toy.
- Topics and aggregator sites keep it growing after the first wave: steady growth of 10–19 stars per day with no single identifiable event, consistent with a long-tail pattern of people finding it via "semrush alternative" searches.
- AI answer visibility (GEO) as a new selling point: the author said on LinkedIn this is the part he's most excited about, and aggregator sites also mention it, making it look like the next generation next to older open-source SEO tools (SerpBear, SEOnaut).
Differentiation, and was it the reason it worked?
- Differentiation: (1) one compose file bundles the full suite, including team seats, white-label reports, alerts, and local/app SEO; (2) "plain language": each issue is explained in one human sentence covering why it matters and how to fix it; (3) 7 languages, including Arabic RTL and Chinese; (4) bring your own keys, no markup, shared cache across users; (5) AGPL with an official cloud version, following the open-core plus hosted-revenue model.
- Is that the reason? Only partly. The part that actually gets repeated is "self-hosted + 26 checks + plain-language explanations," which OpenSEO can also claim. Compared with OpenSEO, it's actually heavier (6 containers), you can't audit your own site without keys (Reposquare's test: "without keys it shows demo data, not your site"), and the cloud version is much more expensive. The stars mainly come from category demand, the Reel distribution, and topic squatting; the product's completeness helps people who click through decide to star it, but it isn't why people came in the first place.
How visible it is
Circulating in small circles: one Instagram repost Reel with 81k views, a batch of AI aggregator and review sites, and one hands-on comparison article. Not on HN, no Reddit threads, zero posts in the X corpus, and no media coverage. The cloud version has public pricing but no public revenue, and no TrustMRR record; rankme.fast isn't in the Tranco top 1M.
Copycats
Nothing that explicitly copies RankMeFast. Same-topic repos that appeared around the same time (springvoiceswell/semrush-ai-tool at 366 stars, 09-25; duckduckdeploy/openseo at 49 stars, 10-15) look more like they're chasing the "Semrush alternative" keyword than copying it.
What others can borrow
- When you see a validated open-source category, build a fuller version that teams and agencies can use, then add a hosted tier to charge for.
- Squat GitHub topics your rival doesn't use: check which topics the leader uses, then grab the remaining hot terms.
- Buy one repost Reel in the first two weeks of launch (gittrend's current rate is $20), and make sure the README's first screen can be cut into a 28-second clip.
- Build a fake backend for each data source so people can click through the whole product without paying, removing the "apply for a key first" barrier.
- Localize the docs; that alone brings in GitCode mirrors and indexing from non-English aggregator sites.
Still unclear
- Whether the gittrend.io Reel was paid or a natural pick.
- Where the 10-05 single-day +182 came from (possibly GitHub's TypeScript Trending list, gittrend's Story, or another post that wasn't indexed).
- Whether the stars are real: stargazer lists aren't accessible, so account quality can't be checked. Many forks come from new auto-named accounts (like xxx-design or xxx-hash), but that proves nothing either way.
- Whether the cloud version has any paying users.